We leverage our data access (Owkin Loop), our data science team (Owkin Lab), and our catalog of models to deliver breakthrough insights. Our offering is structured around 3 products, that can be combined into a full-stack offering or accessed one by one: 

1. Data Connect: connect to the best curated research datasets in the world

2. Models: license predictive models of data enrichment, outcome prediction, and patient identification

3. Full-Stack Services: from data access to bespoke models and consulting services

1. Data Connect by Therapeutic Area

Fit-for-AI data is hard to find. To solve real-world problems, data must be multimodal (raw data such as genomics, imaging, pathology), longitudinal (no missing data points, curated and annotated by the best specialist), and multicentric to gather heterogeneity of patient populations, prescription habits, and data capture.

Owkin unlocks fit-for-AI data by connecting research cohorts around specific therapeutic areas, mutations, or pathways.

Examples of our therapeutic access loops include: NSCLC, Breast Cancer, EGFR, and NTRK.


Owkin Loop

2. Models

Owkin Lab data scientists have collaborated with researchers in the Loop network to build predictive models for data enrichment, outcome prediction, and patient subgroup selection. The Owkin model catalog consists of 30 live models, validated and packaged, with 40 more in the pipeline. These models can be licensed and accessed through Owkin Studio or an API.

Owkin Lab also collaborates with pharma partners to build bespoke models. These models can be trained directly using proprietary pharma data and improved by training and validating on Owkin Loop data in a federated context.

Owkin Models

3. Full-Stack Services

Owkin leverages data access and AI expertise to deliver end-to-end services, that can be deployed across pharma drug development.

Translational Research & Early Stage Development

    • Multi-modal predictive biomarkers and signatures
    • Improve Phase 1 evaluation with counterfactual outcome predictive models

Late Stage Development

    • High-value subgroups for patient selection, stratification, and adaptive trial design
    • Improved endpoints and surrogates for increased objectivity and efficient trial design
    • Enhanced trial evaluation with covariate adjustment 

Market Access & Product Strategy

    • Data-driven product differentiation
    • Companion diagnostic to increase target population
Get in touch